Project Summary/Abstract Our long-term goal is to ensure that clinical decision support (CDS) targeting nurses within the multidisciplinary longitudinal care plan provides efficient and effective support for evidence based nursing care decisions that improve patient outcomes. Evidence points to a tremendous gap between hospitalized end-of-life (EoL) patients' desire for comfort and dignified death and the care they receive. Nurses, who provide the majority of hands-on care for hospitalized patients, are often ill-prepared to provide patient-centered EoL care. Patient specific evidence delivered at the point-of-care to nurses at the right time and in the right format has the potential to dramatically improve patient outcomes. In our team's foundational research, we iteratively built interactive CDS prototypes and demonstrated the feasibility of a pre-clinical (simulation based) randomized controlled trial (RCT) with functional interactive CDS prototypes (text, text+table; text+graph, control) with 60 nurses to compare groups for effects on patient outcomes. The findings showed significant positive impact of all three CDS formats on plan of care decisions associated with improved outcomes for EoL patients. We also found that the nurse's decision time varied with the nurse's graph literacy (GL) under different CDS formats, indicating that the optimal CDS format for a nurse might depend on their GL level. These findings have important implications for translation of CDS interventions into clinical care. A crucial step toward confirming these findings is to fully test the relationship between GL and optimal CDS format in an adequately powered clinical (simulation based) RCT with a nationally representative sample of 220 registered nurse subjects. The testing strategy is innovative and significant since it allows the generalization of findings to systems that comply with national terminology and care plan standards and avoids the unintended consequences that occur when ill-conceived CDS is implemented into live electronic health record systems (EHRs) prematurely. We now propose the following: Aim 1. Compare the four CDS groups (text, text+tables, text+graphs, tailored) for effects on CDS decision time and patient outcomes. We hypothesize that the tailored CDS group will have faster decision time (primary) and better patient outcomes (secondary) than the other CDS groups. Aim 2. Examine associations of other nurse characteristics (e.g., numeracy, format preference, demographics [education, experience]) with CDS decision time and patient outcomes by CDS formats. We hypothesize that (a) higher numeracy is associated with faster decision time and a better patient outcome under text+table and text+graph, (b) alignment between assigned format and nurse preference is associated with faster decision time and better patient outcomes. Findings will enable improved CDS tailoring based on more refined models that predict the RN's decision time and patient outcome under different CDS formats.